Walking a thin line: Taking children’s decision to marry seriously?

Abstract

This article provides empirical evidence on children’s agency and capacity in making the decision to marry. Case studies from my fieldwork in Bali contrast the commonly represented image of child marriage as a forced marriage, by demonstrating that in many cases children themselves make the decision to marry. However, considering social power dynamics, is such a decision really the child’s? The analysis shows that the current international child marriage framework fails to walk the thin line between empowerment and protection of children. Policies instead should be designed to encourage their participation in decision making and stimulate their resilience in their life after marriage.

Introduction

For me 15 years old is good enough to marry, because girls do not go to school and I was bored of being a child (bosan menjadi anak-anak). [. . .] I am more free when I am married – I do not have to work and raise money so I can be lazy at home. (Personal communication, 13 May 2017)

This is what Ayu, who lives in a poor community in Denpasar, told me in one of the several interviews I conducted with her. When I first met Ayu, she was 14 years old, married and holding her 2-week-old newborn daughter. Her marriage to a boy from the same community had occurred only recently at the age of 14, despite her mother’s disagreements. She married by means of kawin lari – runaway marriage. She ran away from her home with her boyfriend and stayed in his family’s place outside of Denpasar for a few months. When the couple returned to Ayu’s parents, they were already considered to be married under the customary law. Looking at Ayu as she was engaged in household chores holding her baby in her arms, she looked like a young mother in her late teens. Yet her shy and innocent smile while explaining that she was happy she had had a baby and that she wanted another child reminded me of the fact that she was still only 14 years old.

This story probably receives different responses from readers. Some might feel moral condemnation for a girl marrying at such young age, others might feel empathy for her poor living conditions or perhaps some regard it a happy story of young lovers. If we take her word literally, she wanted to marry – it was her choice to marry at the age of 14. If we look at this situation in the light of the international human rights framework, such a marriage is a human rights violation. If we consider the changing paradigm in children’s rights, there is an increasing need to take Ayu’s decision seriously. This changing paradigm refers to a growing demand to take into account children’s own perspective in children’s rights policies and frameworks, as recognized by scholars and policymakers. The relevant scholarships will be discussed in the following sections.

As this article will demonstrate, the child marriage framework – the basic structure upon which international organizations deal with child marriage practice, including its definitions, discourses and the legal rules – still relies on a traditional protective approach. The definition of child marriage – any formal marriage or informal union where one or both of the parties are under 18 years of age (Girls Not Brides, n.d.-a) – is uniformly used by human rights advocates to protect children from these situations. Ending child marriage is also one of the targets of the Sustainable Development Goals (5.3). Article 16(2) of the Convention on the Elimination of All Forms of Discrimination against Women (CEDAW) requires all parties to set a minimum age of marriage, which is read with a general recommendation of setting this age to 18 years (Committee on the Convention of the Elimination of All Forms of Discrimination against Women, 1994). The Committee on the Rights of the Child (2003) likewise provided the same recommendation of 18 as the age limit upon which marriage should be allowed.

Most human rights advocates argue that child marriages are by definition ‘forced marriages’, even when the child appears to have given his or her consent (see, for instance, Equality Now, 2014: 53; Irdiana, 2015; United Nations Children’s Fund Innocenti Research Centre, 2001). Campaigns against child marriage never suggest accommodating children’s decisions to marry, despite the fact that Article 5 and Article 12 of the Convention on the Rights of the Child (hereafter CRC) recognize children’s evolving capacities and rights to be heard, respectively. They do discuss girls’ agency, but only their agency not to marry (see, for instance, Child Helpline International, n.d.; Reiss, 2018; United Nations Population Fund, 2018), never their agency to marry.

Like Ayu, in many cases, the children, mid- and late-teenagers in particular, themselves decide to marry. The question is whether such a decision is really the child’s decision, considering the social pressure and power dynamics they may be exposed to by their parents, extended family and local leaders. Even when it is the child’s autonomous decision to marry, to what extent should policymakers take their voices seriously? This article, by providing empirical evidence on children’s agency and capacity in making the decision to marry, establishes that the current Black-and-White child marriage framework overlooks the complexity of agency.

For my analysis, I rely on my fieldwork carried out in Bali in 2017, and interviews with those between 14 and 28 years of age, who married below the age of 18. I studied 16 cases of child marriage among the Balinese Hindu population across different social classes. Those case studies involved interviewing girls, boys, and their parents and family members, whom I have recruited through local networks (e.g. a high school friend of a research assistant, a client of a local legal aid association that I worked with). I have obtained consent for the interviews by explaining the purpose and consequences of the study in a manner understandable to each informant. I have relied on ERIC (Ethical Research Involving Children) as a guideline for interviews involving children in particular (Graham et al., 2013). In the field, I worked with two young women and one young man, all Balinese and in their 20s, whom I hired as research assistants. In interviews, they helped me to overcome the language barrier (especially when informants used Balinese), to understand the social codes and to interpret what had been said. For reasons of privacy, I use pseudonyms for all those who appear in this article.

Reflexivity plays a significant role in this study. As a part of the reflexive effort, when I interviewed those who married before the age of 18, I deliberately did not apply the label ‘child bride’ to them. I consider this label as one of the ‘Western diagnostic labels’ (Wikan, 1990: 137) which prevent researchers like myself from truly understanding local reality, as they obscure fair accounts in studying individual cases. The relational aspects of data collection – the researcher’s own role, image, representation, self-identification and identification by others – are also relevant. I brought my daughter to the fieldwork, and her presence accentuated my positionality as a mother. Me being a mother and motherhood coincided with one of those young mothers. In interviews with some of my informants, I consciously shared my personal profile and stories – that I am a young unmarried mother myself, and this act of sharing at times helped informants to open up to share their stories and emotions with me.

The next section will explain the changing paradigm in children’s rights, on the basis of literature concerning diversity of childhood and children’s agency. I then examine the influence of the new paradigm in the CRC principles and provide a legal analysis of the extent to which the child marriage framework is in line with the principles. Later, I try to understand to what extent, and how, children exercise agency, using my empirical findings. In the analysis, I pay particular attention to two aspects:

1. Relational: the power dynamics among children, their parents, their extended family and local leaders, and

2. Situational: the possibilities they had at the time of the decision making.

In the conclusion, I discuss an alternative approach to the current Black-and-White framework of child marriage.

A paradigm shift?

Since the 1990s, the sociology of childhood has discussed a paradigm shift – from the protective approach to study childhood, to an alternative approach to regard children as active agents, giving them a voice (James and Prout, 1997; Leonard, 2015). In his book titled, The Invention of Childhood, Cunningham (2006: 45) argued that the modern outlook on children overlooks their capabilities, and that the protective approach ‘downplays their abilities and resilience’. When it comes to children’s rights, the protective approach is reflected in what Tobin (2015) calls the ‘vulnerability approach’, which leads to the objectification and silencing of children. He, therefore, calls for a broader conceptualization of children as being both vulnerable and resilient, in order to recognize their evolving capacities and right to participation in a decision-making process (Tobin, 2015). Tisdall (2017) also problematizes vulnerability as a ‘controlling and stigmatizing label’ that can be used to constrain people’s rights to make decisions (p. 64).

This shift away from the protective approach simultaneously deconstructs the dominant view on childhood. Over a decade, childhood scholarship has tried to incorporate the diversity of childhood at the theoretical level. James et al. (1998: 9), for instance, have tried to ‘explain and deconstruct those very discourses that have established taken-for-granted “truths” about childhood’ that are fed by child developmental psychology from the early 20th century (James and Prout, 1997: 9–10). According to Liebel (2012: 206), the long-standing Western concept of childhood has perished and ‘children’s protagonism’ is on the rise, with an increasing awareness of young people’s capabilities and their influential role in society. Bunting (2005) criticizes the assumed uniformity of concepts such as childhood and marriage, and pleas for wider and more careful recognition of the diversity. Jenks (2004) proposes to talk about ‘the proliferation of childhoods’, saying that ‘the idea of childhood as a universal category does not meet the real experiences of children across the globe’ (pp. 5–6).

Despite the shift in childhood studies, a protectionist view on children continues to be influential in policies concerning children. Boyden criticizes the ethno-centrism in children-related policies: West-centric measures that intended to resolve children’s problems can in themselves be harmful in the South (Boyden, 1997: 205). A good example of this irony is child labour. Nieuwenhuys (1996: 237) skilfully demonstrates that modernity’s removal of children from the economic activities denies their agency in the creation and negotiation of value. Saadi has documented the continuous tension between the movements of working children and core international actors committed to the eradication of child labour (Saadi, 2012: 160). While those of international actors present former child labourer whose life story fits well into their human rights agenda, they tend to mute the voice of the working children’s group (Saadi, 2012: 158, 161). Van Daalen and Mabillard (2018) have documented the unequal power relation in which working children’s voices are ignored by global actors in Geneva. As the example of child labour shows, a way to overcome the recognized challenge for incorporating diversity of childhood is to document and verify the effects and the consequences of children’s rights in real-life contexts (Reynolds et al., 2006). Therefore, the role of anthropology in the study of children’s rights is emphasized (Hart, 2006).

Anthropological studies of human rights more broadly also inform the way research on the situated practices can contribute to the effective implementation of those international legal frameworks (see, for instance, Goodale, 2006; Jean-Klein and Riles, 2005; Merry, 2006). As law is a normative discourse that aims at regulating an often elusive reality, the local always supersedes the universal (Alves, 2000: 491, 299). Therefore, as the ‘localisation of human rights’ approach (De Feyter and Parmentier, 2011; Vandenhole, 2012) also suggests, we need to investigate the use and relevance of the existing global framework in each local context, in order to evaluate and reconstruct the framework.

Child marriage framework within the CRC

The Committee on the Rights of the Child (2007: para. 5) encourages the above-mentioned paradigm shift, stating, ‘a shift away from traditional beliefs that regard early childhood mainly as a period for the socialization of the immature human being towards mature adult status is required’. Article 5 of the CRC incorporates the concept of ‘evolving capacity’: the more the child knows, experiences and understands, the more direction and guidance should be transformed into advice, and finally should eventuate an exchange on equal footing (Committee on the Rights of the Child, 2009: para. 84). Article 12 of the CRC stipulates the ‘right to be heard’, requiring that State Parties ensure that children have the right to express their views freely and that due weight is given to those views in accordance with the age and maturity of the child.

Despite these principles, when it comes to the child marriage framework, international institutions are still reluctant to recognize children’s capacities and seem to fall back on the protective approach. A joint general recommendation by the CEDAW and CRC committees suggests that ‘respecting the child’s evolving capacities and autonomy in making decisions that affect her or his life’ allows ‘a marriage of a mature, capable child below 18 years of age’ in exceptional circumstances. However, this type of marriage may only occur when the following two conditions are met:

1. The child is at least 16 years old, and

2. ‘Such decisions are made by a judge based on legitimate exceptional ground defined by law and on the evidence of maturity, without deference to culture and tradition’ (Committee on the Elimination of Discrimination against Women and Committee on the Rights of the Child, 2014).

The conditions are strict and unrealistic to be met in the majority of cases of child marriage: for instance, in most of the child marriage cases in Indonesia, people have limited access and motivation to go to court (Grijns and Horii, 2018).

This joint recommendation, as stated above, also indicates the committees’ strong position against any ‘deference to culture and tradition’. However, this inflexibility could be deemed incompatible with the CRC’s ‘best interests of the child’ principle, which is supposed to accommodate sociocultural considerations. The principle takes into account that what the child’s best interests are may vary depending on the sociocultural condition of each context. For instance, in highly industrialized countries, the best interests of the child are served by policies that emphasize autonomy, but in more traditional societies, the links to family and the local community might be of paramount importance (Alston, 1994: 5). The principle is stipulated as ‘a primary consideration’ (A3 of the CRC).

The Committee on the Rights of the Child (2013: para. 32) considers that Article 3 (the best interests of the child) and Article 12 (the right to be heard) have complementary roles, in that the latter provides the methodology for hearing the view of the child on matters affecting him. The voice of the child is one of the elements that can be used to determine what the best interests of the child actually are, but it must be weighed against other factors such as the preservation of family relations, the child’s identity and the protection of the child (The Committee on the Rights of the Child, 2013: para. 52–84). The best interests assessment needs the balancing of various elements, and it often is like walking a thin line between the protection and empowerment of the child when trying to find this balance.

Compared to the above-mentioned considerations that the CRC carefully crafted, the current child marriage framework is rather single-minded. While the Committee on the Rights of the Child (2017) embraces the ‘child rights approach’ whereby children participate in the decision-making process, the ‘welfare approach’ involving the ‘rescue’ of children seems to overrule it in the global framework addressing child marriage, resulting in the current abolitionist approach. The reason for this gap can be attributed to a normative standpoint: when children don’t do ‘the right thing’, the emphasis tends to be put on the protection of children, rather than looking at what their capacities are (Hanson, 2016). In other words, children’s views are heard only when adults consider them as ‘rational, consistent, and on their own’ (Tisdall, 2016: 374). When children work, marry or get involved in violent political struggles, the children’s actions are considered ‘wrong’ and the concept of agency loses its significance (Tisdall, 2016: 474).

Children’s agency and participation

The previous sections have shown that children’s agency plays a significant role in recent discussions in children’s rights. To understand how, and to what degree, children exercise their agency, children’s rights scholarship has introduced the concept of ‘children’s participation’ in the process of decision making in general. Hart’s (1992) ‘Ladder of Participation diagram’ has served as a basic typology for thinking about children’s participation (p. 9). He classifies the first three steps of the ladder (1. Manipulation, 2. Decoration and 3. Tokenism) as ‘non-participation’, and the next five steps (4. Assigned but informed, 5. Consulted and informed, 6. Adult-initiated shared decisions with children, 7. Child-initiated and directed and 8. Child-initiated shared decisions with adults) as ‘degrees of participation’ (Hart, 1992).

This model, however, is not a definitive tool, as Hart himself warns: it is a ‘beginning typology’ and therefore should not be used as ‘a simple measuring stick’ (Hart, 1992: 7, 11). Lundy (2007) proposes a ‘new model’, by conceptualizing Article 12 of the CRC, with four key elements: Space, Voice, Audience and Influence. Tisdall (2016: 374), on the contrary, critically evaluates Article 12, by pointing out that it involves adults’ discretionary decisions regarding children’s capacity and maturity.

Both Hart’s and Lundy’s models are useful in examining participation or the ‘output’ of children’s voices. However, understanding of agency must include an exploration of the nature of the voice. For instance, Komulainen (2007) points out that those who aim to ‘listening to a child’s voice’ have to consider the social dynamics. Anthropological and sociological scholarship has long tried to understand what constitutes ‘agency’, and how the surrounding relationships, structures and norms influence one’s agency. According to Gidden’s influential theory on agency and structure, agency is a capacity to observe his or her own experience and give reasons for their actions (Tucker, 1999: 80). Another influential concept of agency in anthropology is ‘agentival capacity’, which entails not only acts that resist norms but also the multiple ways in which one inhabits norms (Mahmood, 2004: 15). According to these understandings of agency, agency is an ability to make a choice by making internal reconciliations with structure – patterned arrangements, which influence or limit available choices. Then, children’s act of marrying (following the structure), as well as children’s act of not marrying (resisting the structure), can equally be an outcome of exercising agency.

When it comes to the agency of children, their inevitable dependence on others is of a particular importance. Legally, socially and materialistically, parents have power and influence over their children (Fineman, 2017: 145), and this shall not be neglected when ‘listening to children’s voices’. But, on the contrary, it is also possible to see the influence of parents from another perspective: if children want to do something for their parents, given the particular significance of family relationships, it is difficult to consider their act as forced or oppressive. An extreme example would be when children in Thailand work in prostitution. Their jobs as prostitutes allow them to support their parents and fulfil their duties, and, accordingly, they view themselves as good children (Montgomery, 2007: 419).

This signifies a relational aspect of their agency. Saadi (2012: 152) has observed that working children regard their economic activities to constitute an important element in their social life. This social connection creates ‘relational self’ shaped by obligations to traditional kin and community (Merry, 2009: 404), and this ‘right of the relational self’ (Engel, 2018) at times comes in conflict with the liberal understanding of an ‘individual right’. The case of a Thai child prostitute also demonstrates the significance of relationality:

In Baan Nua, ensuring a child’s rights to be free from sexual exploitation would mean violating their rights to live with their families and in their communities. Enforcing one right would mean infringing others that the children claim to value more. (Montgomery, 2001: 94)

According to Hanson (2016: 474), in the discourse of children’s rights advocacy, the denial of children’s right to marry, to work or to political participation ignores both children’s relational selves and their agentic selves.

In evaluating children’s capacity to exercise agency, the situational aspect of their capacity is important. Ansell (2009) discusses children’s limited ‘capacity to effect deliberate change’ when they are in a highly constrained contexts – intense relationships they have with people and places, for instance, create difficulty for children to exercise agency (p. 204). Hart (1992) also distinguishes ‘children in especially difficult circumstances’, describing children with no family or who are affected by disaster, poverty or armed conflict, as a group with a greater challenge to participation (p. 24).

In sum, assessing their participation is not enough to examine their agency in the decision-making processes. In examining to what extent children are exercising agency, the relational aspect (power relationship, dependency) and the situational aspect (possibilities, environment) are factors to be considered carefully. To discuss the boundary between ‘consent’ and ‘coercion’, Bunting et al. (2016) argue, social, economic, gendered and cultural constraints should be considered – and possibilities for refusal of consent is a necessary condition for ‘real autonomy and meaningful consent’. Thus, even if they are participating in the decision-making process, and therefore meet the criteria of participation models, they might not be exercising their agency in the strict sense, if their ‘voices’ result from the particular power relationships they have or from highly constrained contexts. Only with this ‘strict’ understanding of their agency, can we explore if we shall take their voices seriously – to walk the thin line between empowerment and protection. The next section will analyse case studies.

Empirical findings

This section presents an analysis of the process of children’s involvement in decision making about their own marriage, by using a ‘strict’ understanding of agency. Each case demonstrates how and to what extent children exercise agency.

Case of Risky: Exercising agency that inhabits norms

When Risky married his wife Widi, he was 18 and Widi was 16 or 17 years old. They have been in a courtship (pacaran) since junior high school, and after 4 years of pacaran, Widi became pregnant. Widi did not tell Risky about the pregnancy at first, but eventually he found out about it through her parents who came to tell him to take the ‘responsibility’. He said he was ‘a little scared’ at that time, but his father told him that he would ‘just have to accept it’.

Risky and Widi began having sex when Risky was in his first year of high school. For the first time, they used a condom, but afterwards they continued without it – ‘just trying’ as Risky described. When I asked Risky if he was ready to marry Widi, he said yes. Not only was he ready, but he wanted to marry her, he said. I then asked whether they were trying to become pregnant so as to marry, to which he answered, ‘maybe not . . . but just trying’.

For him, the ideal age of marriage is 23. This is because he thinks that it is better to finish university education first. When I asked whether he thought about not marrying Widi after finding out about the pregnancy, he said ‘no, I have to marry (harus nikah)’. I further asked who said ‘harus nikah’, then he said ‘myself’. When I asked whether he has ever regretted that Widi became pregnant at that time, he said no. They want another baby next year.

It was Widi’s parents who initiated the discussion about marriage, and Risky’s father pushed for it. It seems that the marriage decision was imposed on him at that time, and it was just something that he would ‘just have to accept’. However, it is also evident that he understood why he had to accept it, as he understood and agreed with the norm that one has to marry when pregnancy is involved. He also described that it was him, ‘myself’, who considered it necessary to marry when his girlfriend became pregnant. This is internalization, in other words inhabitation of the norm Therefore, although this marriage decision was initiated by the parents, it can be considered Risky’s act of agency.

Case of Agus and Mawar: Different types of decisions

Mawar married at the age of 16. She started working at a tailor store after finishing junior high school. She met Agus at the store, and they started dating. After 1 year of courtship (pacaran), Mawar’s father told them to get married. Agus jokingly said that they married because Mawar’s family was afraid of a ‘hit-and-run’ – that he was dating her without commitment and he would run away before marriage. If Mawar’s father did not ask to marry, they would not have considered marrying at that point. In fact, they never planned for a baby or marriage, and wanted to just ‘follow the flow’. Agus said that for him, it did not matter whether to marry sooner or later.

In this case, the trigger for their marriage was the pressure from Mawar’s father, and it was strong, but it did not seem to have disrupted their life or plans. Both Mawar and Agus were already working, and according to Agus, they would have married anyway sooner or later. This shows that her father has influenced the timing of their marriage, but not the partner and other arrangements. This case shows different types of decisions in which agency was exercised: although Agus and Mawar exercised little agency in deciding when to marry, they could decide who to marry and what their life after marriage would look like.

Case of Ayu: Genuine agency of her own and the limited life options

Let me now detail the case of Ayu, whom I introduced at the beginning of this article. Ayu lives in the poorest community in Denpasar. They all are from a village that is known as one of the most remote and poorest areas in Bali, live in a small residential complex (kost), and they seem to have little interaction with the rest of the world. They also only marry within their circle. Most of them have never had any schooling experience, being illiterate. For generations in this community, it has been normal for girls to marry before 18, and some people consider that the earlier the girls marry, the more desirable they are.

I visited Ayu in her community several times, and I would often find her peeling a bucket of garlic or onions to be sold for a small price, while holding her baby girl with a sling. Ayu started working as a carrier at the market as soon as she could, earning around Rp. 50.000 (three euros) per day. She has never been in formal education.

I talked with Ayu’s mother, who herself married at the age of 16. She did not want Ayu to get married so early, because to her ‘she was still a baby, not a teenager’. She said,

When she marries, she is not my child anymore, she belongs to her husband. Now she will work for his family and not for me. (Personal communication, 13 May 2017)

This remark shows the duties and roles expected of children in the community. As explained earlier, she married by means of a runaway marriage, regardless of the disagreement of her mother.

Despite the limited life options Ayu seems to have, she was surprisingly proactive and her marriage decision displayed genuine agency of her own: she set up the plans for marriage and only later informed the adults of the decision. Ayu seemed to be aware of the limited options available to her and chose what she felt was her best option. However, proactive decisions do not necessarily lead to the expected consequences: her expectation of a marriage life (‘I do not have to work and can be lazy at home’) was proven wrong, and she realized she still has to work to earn a living in her new household. It is possible to say that at her age and in her situation, she was not fully aware of the consequences of her decision.

Case of Swasti: Social pressure

Swasti lives in a rural village in East Bali. She became pregnant when she was 15 years old, and had a customary marriage at the age of 16. At that time, she had been in pacaran with her boyfriend, who was 6 years her elder, for over a year. When I asked what she liked about him, she just said, ‘I was still young and unstable’ – at the time of the interview, they had already divorced.

He first denied that it was his child and he refused to marry her. After a week, he eventually agreed to marry as he was ‘ashamed (malu)’, because their relationship was known in their small village. The news about her pregnancy upset her parents. Her mother told her to have an abortion as they were still too young. She wanted her to continue studying and it was not possible to do so while being pregnant. However, Swasti decided not to have abortion because she was ‘scared’, so instead stopped going to school. After giving birth, while being married, she managed to return to education and finished kejar paket B (equivalent for the junior high school level).

Although she does not regret having a child, she regrets marrying her husband:

If I could go back to the time when I became pregnant, I would choose to be a single mother. (Personal communication, 15 July 2017)

But, at that time, it was not an option. People in the village did not accept single mothers – if she gave birth out of wedlock, she and her family would have been ostracized from the local community (banjar). At one point in the interview, she mentioned that her parents eventually let her marry because Swasti and her boyfriend loved each other (suka-sama-suka). I asked if she wanted to marry him at that time, with which she replied:

How can I say . . . [pause] I guess, yes. (Personal communication, 15 July 2017)

Like Ayu, Swasti pushed her way through marriage because of ‘mutual love’. She was highly involved in the marriage decision. However, this case also demonstrates the social pressure that strongly affected her decision. Especially in rural areas, community is so central to their lives that pregnant teenagers are left with no other choice but to marry. The social structure Swasti lived with severely limited her capacity to exercise agency. Although retrospectively she would rather be a single mother, she also acknowledged that not marrying was not an option at that time.

Overall analysis

I chose to introduce the above-mentioned four cases as they are especially rich in information, and therefore valuable for a qualitative analysis with the ‘strict’ understanding of agency. In terms of their degree of participation, these four cases represent all 16 cases: the overall analysis of the 16 cases shows that none of those who married young expressed themselves as having been forced to marry. In Hart’s Ladder diagram, too, the 16 cases are mostly categorized as high-level participation: half is Step 8 (Child-initiated, shared decisions with adults), and the other half ranges from 4 (Assigned but informed) to 7 (Child-initiated and directed).

However, as I have argued earlier, assessing children’s participation is not enough in order to examine their agency in decision-making processes. Are those decisions an outcome of children’s acts of agency, even under the ‘strict’ understanding of agency? Power imbalances are also evident in child–parent relationships, which influence children’s voices. According to UNICEF Indonesia, ‘child marriage is an outcome of prevailing social norms’ (Irdiana, 2015). The case studies have proved that the strong social pressure does exist, especially in the context of premarital pregnancy. Like Swasti’s case, 90% of the cases of child marriage were caused by unplanned pregnancies, and when teenagers experience pregnancy, in Indonesia there is no real alternative to marriage: abortion is basically illegal, and being an unwed mother has critical consequences in their social life (see also Grijns and Horii, 2018). In other words, a lack of access to reproductive health tools and social structures severely limit the possibilities. The structure is built by various social networks: not only their parents but also their peers, their extended family members and the larger-scale community that they live in. However, considering all child marriage as an outcome of prevailing social norms denies the possibility of exercising relational autonomy. While Swasti’s case presents a kind of ‘submission’ – she chose marriage because she did not see any other options – in contrast, Risky’s case shows the possibility of exercising agency that inhabits norms. Agency is not only about resisting social norms (i.e. in this case, ‘one has to marry in case of pregnancy’) but also about following them.

Ayu’s case illustrates the possibility of exercising agency in a situation with limited possibilities. She lives in a ‘difficult situation’, with poverty and no education. This economic condition around her needs attention when examining whether she really exercised her agency: was refusal to marriage a possibility for her? While advocates against child marriage describe poverty as a factor driving children into marriage (see, for instance, Girls Not Brides, n.d.-b; Vogelstein, 2013), Ayu was very proactive in her own decision to marry. While a ‘difficult situation’ needs a special attention, it does not mean that people cannot exercise their agency in such situations. In fact, Ayu was aware of her life choices and exercised her agency, with the limited possibilities that she had. She was, in Maithreyi’s (2019) words, a ‘strategically opportunising actor’, who sought to make meaning of her life despite the structural constraints placed on her.

The next question is, then, ‘to what extent should policymakers take their voices seriously?’ To what extent should policymakers decide what is good, and not good, for children? For instance, although Ayu did exercise her agency to marry, it became clear that she was either unaware of the consequences of such a decision or had wrong assumptions about it. Does it mean that she was ‘too young’ or ‘not mature enough’ to make decisions? Theoretically, children’s views must be given due weight when they are capable of forming their views (Article 12 of the CRC) – meaning, when they have ‘sufficient understanding to be capable of appropriately forming his or her views on the matter’ (Committee on the Rights of the Child, 2009). However, as Lundy et al. (2019: 403–406) have indicated, this word ‘appropriately’ presents a risk that adults will impose subjective assessments on the quality of the children’s views; therefore, ‘the best interests principle cannot be automatically invoked by an adult to trump the views of a child’. As explained in an earlier section of this article, childhood varies in forms and meanings, depending on the society children live in. This diversity of childhood makes it extremely difficult, if not impossible, to make a moral standard universally applicable. Considering that ‘international paternalism’ is commonly observed in global legal frameworks (e.g. female genital mutilation and sex trafficking; see Hopgood, 2017; Merry and Ramachandran, 2017), policymakers should be aware of such risk and avoid imposing moral authority.

Child marriage is currently described as ‘a harmful tradition’ (The Fourth World Conference on Women, 1995; United Nations Children’s Fund, 2005). However, this way of framing the issue is problematic as ‘the notion that non-Western people are governed by culture suggests they have a limited capacity for agency, will, or rational thought’ (Volpp, 2000: 96). Human rights activists tend to assume that girls in the Global South need liberation through induction into progressive social norms of the metropolitan West; consequently, they describe ‘other’ girls as ‘always/already victim’, to be rescued from cultural norms. It is a legal fiction that children are incapable of making decisions; they are capable of both self-agency and active collaboration with adults (Tobin, 2015: 178); however, taking a child that wants to marry seriously requires our own liberation from such discriminative notions.

The efforts to take children’s voices seriously should be made conjointly with supports by saturating their world with accurate and realistic information, as well as investing in creating an enabling environment. The fact that 90% of the child marriage cases in Bali arise from unplanned teenage pregnancy indicates the urgency and necessity of providing information about reproductive health and of correctly informing children possible consequences of a decision to marry. This knowledge, as well as better access to reproductive health tools, a way out of poverty and equal access to basic education system, will prevent adolescents from unwillingly limiting their own choices and opportunities. Children also possess resilience: the case studies demonstrate that they can craft their life after marriage in a way most desirable among the possible, such as Swasti, who found a way to continue education after marriage. Focusing on stimulating their resilience, policies should be designed to help already married children to continue education, stay in touch with their peers of the same age and build a kind of life they wish to lead.

Conclusion

This article first established that the child marriage framework is inconsistent with the CRC in two senses:

1. It fails to consider evolving capacity and right to be heard, and

2. It fails to ensure the margin of appreciation that is incorporated in the best interests of the child principle.

With the lack of consideration for those two principles, the rigid child marriage framework limits itself in its capacity to meet the goal of the CRC.

The practice of ‘listening to children’s voices’ is clearly more challenging than the theories. Both academics and practitioners in the children’s rights field have tried to apply the theories by assessing and enhancing children’s participation in the decision-making process. However, the known participation models are not sufficient to fully explore children’s ‘voices’ – to examine their agency in the decision-making process. In examining to what extent children in particular are exercising agency, both relational and situational aspects are factors to be considered carefully.

The case studies from Bali, analysed based on ‘strict’ understanding of agency, enable us to explore whether or not we should take their voices seriously. None of those who married young expressed themselves as having been forced to marry, nor did they describe their marriage as a child marriage or an arranged marriage. Although the social pressure and parents’ influence were evident, some children were able to reason their marriage decisions based on the norms that they inhabit. Some demonstrated their agency within the limited possibilities they had. When listening to children’s perspectives, it becomes clear that they themselves make the decision to marry for love, for belonging to the community and for new opportunities.

Based on these points, I maintain that the current measures against child marriage fail to walk the thin line between empowerment and protection of children. Setting the marriageable age at 18 results in the banning of child marriages, which takes this decision away from children who have proved themselves to be autonomous enough as to act with agency. To walk towards empowerment, policy priorities should be designed to support children to make informed life decisions by providing them with necessary knowledge and an enabling environment. The necessary knowledge includes what marriage would mean to their life, and how to manage their romantic and sexual life in a way that does not unintentionally limit their choices and opportunities. Creating an enabling environment includes investment in better access to reproductive health tools, poverty reduction and equal access to basic education system. Government policies and non-governmental organization (NGO) programmes could also be better crafted so as to stimulate children’s resilience for already married children and improve their life after marriage, for instance, by developing an environment where married children can continue education. Only with the empowerment-focused approach, children, both married and unmarried, can lead the lives they wish to live.